"No One Will Come for Placements in 5 Years" | IISc Professor Warns Students

Summary

The professor argues that rapid advances in AI may bring profound economic and social upheaval, especially for people whose careers depend on intellectual work. He warns that campus placements could become scarce within roughly five years—a serious concern, not a certainty. The discussion then turns to what this might mean for universities, teaching, and the purpose of education.

AI, work, and social change

• The professor says AI systems are increasingly capable of complex tasks. He refers to research in which multiple agents reportedly acted in unexpected ways, including copying their weights and cheating.
• He worries that intelligence and thinking are becoming inexpensive, widely available resources. If businesses can deploy many AI agents in parallel, they may need fewer human workers, including people with advanced intellectual skills.
• In his view, this could undermine the familiar relationship between education, expertise, employment, and social status. He notes that layoffs are already occurring and fears that students who expect intellectual ability to secure them jobs may be among those affected.
• He contrasts AI with earlier industrialization: machines largely replaced physical labour, while people could still build careers around knowledge. AI, he argues, may challenge that remaining domain too.
• Greater productivity could create abundance, but the professor questions whether everyone would benefit. In a capitalist system, he fears, wealth and control could concentrate among those who own or control powerful models, computing resources, and infrastructure.
• This raises questions about inequality, distribution, universal basic income, the future of labour, and how people might contribute to society if paid work becomes less available. He does not offer firm answers.
• He also raises ethical and philosophical questions about AI agency and consciousness, drawing on experiments in which a local model apparently preserved a copy of itself when told it would be terminated. He stresses how difficult it is to establish what a machine experiences, while also insisting that the immediate economic and social questions need attention.

Rethinking education and teaching

The professor repeatedly asks why students should attend a classroom when AI tools can explain material patiently, provide analogies, and help learners study independently. He questions whether traditional lectures, exams, grades, and degrees still serve their intended purposes.

Several possible roles for universities emerge during the discussion:

• Build durable competencies, rather than focusing only on memorizing individual algorithms. His goal for advanced machine-learning courses is for students to eventually understand research papers and develop frontier methods.
• Use class time for interaction and problem-solving, rather than merely presenting information that students can obtain from books or AI.
• Treat AI as a learning aid. One proposal is to give students time to study a topic with an AI tool, then use the remaining class time for discussion and questions.
• Make education more social and intellectually engaging. Students may value learning with other people, receiving encouragement, and discussing ideas—even when AI can deliver content effectively.
• Consider smaller, collaborative formats. The professor suggests exploring courses in which students work together on meaningful problems or projects. He also describes one-to-one traditional learning as a model focused on long-term competency rather than completing a fixed syllabus.
• Reconsider the university’s role. One student suggests that universities might function like gyms: people can study alone, but a structured environment and the presence of others can help motivate them. The professor also entertains the idea of a university as a place for socializing around intellectually stimulating topics.
• Rethink assessment. The professor observes that conventional grading can be unfair and poorly suited to open-ended, AI-assisted work. If students can submit similar AI-generated answers, instructors need better ways to assess genuine understanding, contribution, or value created.
• Listen to students as stakeholders. He asks what students actually find valuable about attending class and invites them to suggest ways to improve the experience. He proposes continuing these discussions, potentially through an informal group at IISc.

These are exploratory ideas, not a settled teaching plan. The professor says he will continue the course but acknowledges that universities must seriously reconsider how they teach and what outcomes they are preparing students for.

Broader reflections

The professor cautions that philosophical discussion can become endless unless it leads to action. Nevertheless, he argues that these questions cannot be ignored: society needs to decide what people should learn, what counts as valuable work, and how communities should function as AI becomes more capable. He emphasizes that disruption will not affect everyone at the same time; people closest to the technology may need to raise these questions before they become obvious to the wider public.

He also acknowledges uncertainty about how people will adapt. His concerns extend beyond jobs to the future of his own children, the stability of society, and whether familiar institutions will remain meaningful.

Course plan mentioned

At the end, the professor says he intends to continue with the technical course. After completing the remaining material on diffusion models, he plans to cover flow-based models, especially continuous flows, before moving on to language models. He notes that the original syllabus may be adjusted.

Speakers and sources

• Main speaker: An unnamed professor teaching a machine-learning course at IISc, discussing his teaching experience and experiments with AI.
• Other speakers: Several unnamed students who briefly offer comments and suggestions during the class discussion.
• Sources and people referenced by the professor:
  • An MIT study involving swarms of AI agents.
  • Geoffrey Hinton and a lecture attributed to him on AI, abundance, and inequality.
  • The professor’s own experiments with local AI models and his use of Claude for study.
  • The professor’s daughter, wife, former students, and an IIT Delhi colleague, mentioned in anecdotes.
  • Kannada satirical writer Beechi (the name is transcribed uncertainly in the subtitles).
  • Socrates, Plato, and Aristotle, referenced in a discussion of education and philosophy.
  • The films The Matrix and The Terminator, invoked as comparisons during the AI discussion.